Vaginal stenosis is a common adverse event following pelvic radiotherapy in women treated for cervical cancer. Apart from the impact on quality of life [30, 31], this may also hinder early diagnosis of tumor recurrences . Therapy is unclear and guidelines are based on few clinical studies and experiences of individual cancer services . In order to add data on the subject, two dependent variables (CTCAE v3.0 scale and vaginal volume) were used to evaluate the effect of topical estrogen, testosterone, and water-based lubricating gel use, as well as daily use of a vaginal dilator, in the prevention of the evolution of stenosis after radiotherapy for cervical cancer.
The incidence of stenosis in the general group was high. At the beginning of the intervention period, 74% of women had some degree of stenosis assessed by the CTCAE scale. At the end of the study, considering only the 142 women who completed follow-up for 12 months, 96% had Grade I/II stenosis. These findings are different from those reported by Kirchheiner et al, who found a 59% crude incidence of vaginal stenosis (43% G1, 15% G2 and 1% G3) after a mean follow-up of 15 months . We believe that this difference is due to the fact that Kirchheiner et al used image-guided adaptive brachytherapy (IGABT) based on repeated volumetric imaging (computed tomography [CT], magnetic resonance imaging [MRI]), unlike our study, where 2D brachytherapy was used. It is known that one of the factors that influences the incidence of vaginal stenosis is high cumulative treatment dose . In image-guided adaptive brachytherapy, it is possible to assess the exact dose administered at each treatment site. With this, it is possible to assess whether the dose limits of each structure are being respected, enabling the assessment of the percentage of vaginal volume receiving radiation. In the present study, women were submitted to 2D brachytherapy, that is, guided only by x-ray image with the anterior and lateral view of the treatment field. Thus, the doses in each structure were calculated at specific points, following the ICRU 38  and it was not possible to calculate the dose received in areas other than these specific points. In a previous publication, we described a 1.74% reduction in diameter and 5.76% in mean length of the vagina in 139 women shortly after the end of radiotherapy for cervical cancer . In the present study, using diameter and length measurements to estimate vaginal volume, we observed that the majority of women (92%) had a reduction in vaginal volume after 12 months of treatment, with a mean decrease of 25.47%.
Despite the rationalization that, by distending the vaginal walls, dilators may prevent adhesion formation in the mucosa and maintain vaginal patency , to date, no study has conclusively demonstrated the benefits of such use . Through the CTCAE v3.0 scale, we observed that, among women who completed 12 months of follow-up, daily vaginal dilator use prevented the evolution (both onset and worsening) of vaginal stenosis. Daily vaginal dilator use appeared to favor the maintenance of sexual activity with less discomfort among these women, preventing the progression from Grade I to Grade II stenosis. However, daily use of a vaginal dilator did not prevent the decrease in mean vaginal volume, which occurred similarly among the four study groups. We emphasize the fact that six women randomized to vaginal dilator use had an increase in or maintenance of vaginal volume, with a trend of not reducing vaginal volume compared to other treatment groups (p = 0.056). Besides that, studies show that adherence to treatment with vaginal dilators may be small . In our study compliance was assessed at follow-up visits by asking the patient if she was able to use the prescribed intervention without difficulty. If problems were identified, a new orientation was performed. Between the follow-up visits, if there were any doubts, women could contact the researchers for clarification. Even so, we observed that two women did not use the dilator correctly during the follow-up period. It is possible that, with greater adherence to treatment, significant benefit with respect to vaginal canal measurements may be achieved.
The use of vaginal dilators is not always tolerated by women with cervical cancer who may use them less frequently than directed or not at all. Therefore, investigation of other methods to prevent vaginal stenosis is essential . Histologically, the vagina is composed of three layers: fibrous, muscular, and stratified squamous epithelium. Estrogen, the main regulating hormone in vaginal physiology, acts mainly on alpha-type estrogen receptors, which have the highest density in the deepest two-thirds of the vaginal canal [36, 37]. Additionally, some authors suggest that estrogen receptors are present in sensory and autonomic neurons of the vagina and vulva . In the presence of adequate serum estrogen concentrations, the epithelium remains thick, well vascularized, and adequately lubricated, facilitating penetration of the vaginal canal. Genitourinary menopause syndrome (GSM) refers to the set of vulvovaginal signs and symptoms resulting from hypoestrogenism, involving changes in the major/minor lips, clitoris, vestibule, vagina, urethra, and bladder . Estrogen therapy promotes vaginal cell growth, cell maturation, lactobacillus recolonization, increases vaginal blood flow, decreases vaginal pH to premenopausal levels, improves thickness, vaginal elasticity, and sexual response [36, 39]. A systematic review comparing vaginally administered estrogen-based preparations for at least 12 weeks in postmenopausal women concluded that estrogen improves the symptoms of genital atrophy in postmenopausal women compared with placebo . An alternative treatment, tested in research protocols, is topical androgen therapy. This may act on specific receptors in the vaginal canal, or on estrogen receptors following peripheral conversion by the aromatase enzyme . Both estrogen and androgen-based topical therapy for the treatment/prevention of post-radiotherapy vaginal stenosis have been poorly investigated to date. In the current analysis, the use of both estrogen and androgen were unable to prevent the progression of vaginal stenosis as assessed by the CTCAE scale and vaginal volume measurement.
As the study groups were not homogeneous in staging and tumor size, with a smaller number of tumors > 3 cm in size and with advanced stages in the group randomized to the lubricating gel group, we chose to perform two multiple analysis models, one for the CTCAE scale (Poisson regression) and one for the variation of the percentage of vaginal volume (linear regression). The only variable independently associated with a worsening of the CTCAE scale was having sexual activity in at least one of the four evaluations during the intervention period. We highlight the fact that this association is not due to a negative effect of sexual intercourse on vaginal health. This negative association may be related to the presence of symptoms during sexual activity such as bleeding and dyspareunia, which would classify women as having Grade II stenosis. That is, sexually active women are more exposed to the risk of complaining of stenosis affecting vaginal function when compared to women who have not had sexual intercourse during the intervention period. We believe that the use of a vaginal dilator was not significant in this regression analysis due to the small sample size.
Treatment with EBRT and brachytherapy, not having vaginal deliveries, and not having sex were factors independently associated with the percentage reduction in vaginal volume. According to previous studies, the association of EBRT and brachytherapy, high dose rate brachytherapy, and high doses of radiation are associated with a higher incidence and greater severity of vaginal stenosis [40, 41]. Traditionally, sexual activity is recommended to prevent vaginal stenosis . This activity may help to distend the vaginal walls, resulting in a smaller reduction in vaginal canal volume. An alternative hypothesis is that sexually active women are easier to examine gynecologically, facilitating the measurement of the vaginal canal, resulting in more reliable measures. Similarly, with women who had had vaginal deliveries and possibly already had larger vaginal canal dimensions, gynecological examination and vaginal measurement would be easier.
This study has several limitations. Women included in the study were heterogeneous as to the type of cancer treatment to which they had been submitted. The group randomized to receive lubricating gel had a larger number of early-stage, smaller tumors. However, we emphasize the fact that tumor extension to the vaginal canal was similar between the four treatment groups. Sample size was calculated considering the prevalence of vaginal stenosis after estrogen and vaginal dilator treatment in previous studies. During the recruitment and intervention period, some factors prevented us from reaching the previously stipulated sample size. These included the high refusal to participate in the study, which increased the time planned for its completion. With the number of subjects who completed the intervention period, we estimate that the power of our sample to assess worsening vaginal stenosis using the CTCAE v3.0 scale was 44.9% and percentage change in vaginal volume was 10.8%. For a power of 80%, we would need 63 women to complete the study in each group to assess the CTCAE scale and 508 women per group to assess vaginal volume, which would be impossible due to the deadline for the end of the research. The CTCAE scale was also used in women who did not have sex before the evaluations. Because this scale takes into account the influence of the adverse event on organ function (dyspareunia interfering with sexual intercourse), this factor may have influenced the results. However, the number of sexually inactive women was similar in the four groups at the four time points evaluated; thus, the frequency of vaginal stenosis assessed by the CTCAE scale could be similarly influenced in the four groups.
Despite the limitations, we believe that the results obtained are valid. We emphasize the fact that the same physician performed all evaluations, both initial and follow-up, eliminating the possibility of interobserver variation. As demonstrated in our results, the lack of significant correlation between the two forms of evaluating vaginal stenosis corroborates the fact that they can be used in a complementary way. The assessment of the percentage change in vaginal volume allowed an objective assessment of the decrease in vaginal volume. The use of the CTCAE scale added the influence on organ function to the classification of severity of the objectively observed stenosis, complementing the assessment by volume. In our clinical experience, decreasing vaginal diameter in women with a small size before treatment results in greater interference with sexual function compared to women with larger diameter. Thus, the same percentage variation in vaginal volume may result in greater or lesser severity of stenosis for women depending on vaginal diameter before radiotherapy begins, highlighting the importance of using a scale that includes sexual function.